Post Go back to editing

ADF4150HV frequency range in datasheet vs ADIsimPLL


I am currently trying to design a single frequency reference at 14.65GHz. I am using the ADIsimPLL software and got to a design, but there seems to be an issue I wanted to ask about. The design was intended for a HMC531 VCO with the divide by 4 (3.6625‬ GHz) output being fed back into the PLL chip. The software predicted the VCO requires a tuning voltage of ~8.2V so I choose the ADF4150HV to avoid needing an amplifier between the PLL chip and the VCO. However, when i was looking at the datasheet for the ADF4150HV (pictured below) the maximum RFin frequency is shown to be 3 GHz. The design ADIsimPLL produced was for a divide by 4 output of 3.6625GHz. When I looked at the datasheet for the ADF4150 it's max RFin is 4 GHz, capable of handing the divided by four output.

Am I missing something important? Is the ADIsimPLL software wrong? Is the datasheet wrong? Any further insight into this would be much appreciated. Thanks.


Bonus question: In the ADF4150 does the max RFin frequency vary depending on input RF power? For example, with the output buffers disabled, as long as the input power is greater than -5 dBm the chip will operate up to 5 GHz? However, if the power drifts below -5 dBm the max inptut frequency will drop to 4 GHz?

  • Hi Sami,

    There seems to be an error in the ADF4150HV ADIsimPLL model where it lists its maximum RFIN frequency as 4.35GHz when it should be limited to 3GHz as per the datasheet. This will be addressed in the next ADIsimPLL version.

    It appears the ADF4150 and ADF4150HV have their RF sensitivity spec'd differently. I expect in practice the ADF4150HV behaves similarly to the ADF4150 in that the maximum frequency of operation for a given RFIN power is greater when the RF output buffer is disabled. But in any case we decided to not put limits in at final test in the ADF4150HV for RF output buffer enabled vs disabled, and to only spec it for the worst case scenario when the RF output buffer is enabled. It was probably done this way to streamline the datasheet spec table. 

    For your extra question yes there needs to be a signal of power at least -5dBm and not exceeding 5dBm to guarantee operation across temp/supply/process up to 5GHz (with RF buffer disabled). These limits correspond to what we test the part for in the production test (with a little guardbanding too of course). If your signal drops below -5dBm in this example there is the chance the part might not lock at 5GHz across temp/supply/process as it's outside of what we have characterised the part for. Generally the min sensitivity reduces as frequency decreases hence you can get away with a lower RFIN signal power at lower frequencies which is why the sensitivity window as per our spec is more forgiving at 4GHz. Figure 4 and 5 in the ADF4150 datasheet should make this more clear.

    With all that said, the ADF4152HV is probably the part you want to use here as it's spec'd for your frequency range.